Overview
Aleph 2 is a video editor that applies targeted visual changes to existing footage using a text prompt, without requiring any reshooting or frame-by-frame retouching. You describe the change you want, and the model propagates it consistently across every frame in the clip. On Picasso IA, this means a creator can upload a 30-second multi-shot sequence, type a description of the new look, and receive a coherent edited video in a single pass. It reads the temporal context of the whole clip, so cuts and camera angle changes do not break the consistency of the edit.
How It Works
- Upload a video between 2 and 30 seconds long (under 16 MB), then write a text prompt describing the visual change you want applied to the footage.
- Aleph 2 reads every frame in context and applies the edit uniformly from the first shot to the last, even when the camera angle cuts mid-clip.
- If you need the edit to match a specific look at a particular moment, attach up to 5 keyframe reference images.
- Assign each keyframe image a timestamp in seconds, or tag it as "first" or "last" to pin it to the start or end of the clip.
- The model blends your reference images into the edit at the positions you set, then generates the final video for you to download.
Frequently Asked Questions
Do I need programming skills or technical knowledge to use this?
No, just open Aleph 2 on Picasso IA, adjust the settings you want, and hit generate.
How long does it take to get results?
Short clips under 10 seconds typically process in under a minute. A full 30-second video with keyframe references may take a few minutes depending on server load. You will see a progress indicator while the model runs.
What video length and file size are supported?
The model accepts videos between 2 and 30 seconds long, with a maximum file size of 16 MB. If your clip is longer or heavier, trim and compress it before uploading. Standard formats like MP4 work reliably.
Can I control where in the video the edit takes effect?
Yes. Attach up to 5 keyframe images and assign each one a timestamp, the first frame, or the last frame. The model uses those anchors to apply your edit with image-level precision at the moments you specify, then interpolates the look across the surrounding frames.
What if the result does not match what I had in mind?
Try making your prompt more specific about the visual outcome you want, for example describing colors, lighting conditions, or surface textures in plain terms. Adding a keyframe image that shows the exact look you are going for usually produces a closer result on the next run.
Is it free to try?
Picasso IA gives new accounts free credits, so you can run Aleph 2 on your own video before committing to a paid plan. Check your account dashboard to see your current balance.
Where can I use the edited videos I generate?
The output is a standard video file you download directly. You can publish it on social platforms, include it in a client project, or import it into any video editing timeline without restrictions from the platform.
